[Opinion] Welcome, GPT!

A Major Turning Point for Media Reform Returns
Though Media Slow to Adopt New Technologies and Systems
Must Avoid Repeating Past Mistakes of Losing Leadership

On the day I first got a cordless phone in my hand, I was delighted, thinking the agony of holding two coins and waiting for my turn at a public phone was finally over. Senior colleagues countered that the much bigger revolution was when we started typing manuscripts on keyboards instead of writing them by hand.

Since then, we journalists have witnessed the emergence of various new technologies, and each time, concerns about the crisis or demise of journalism coexisted with hopeful "trying it out" type articles. At a glance, the day laptops were issued and the day essential tasks like searching, shooting, and recording became possible with just a smartphone were major events that changed the very nature of journalism.

Recently, I have often heard that something comparable or even more shocking than these changes has appeared. "What impact will generative artificial intelligence have on journalism?" After digging through various reports and asking GPT, the summary is as follows.

The main negative impacts are job reduction and issues of reliability. However, even if the author is somewhat enchanted by GPT's greatness, these two risks seem fully surmountable by human effort. Like any professional, journalists spend a significant amount of time on simple repetitive tasks daily and expend great energy on data collection that someone else doing would be welcome. The job replacement effect of AI is both a crisis and, conversely, an opportunity to provide journalists with conditions to focus more on their core work.

Regarding information reliability, the International Federation of Journalists (IFJ) report’s advice that "there must always be a human alongside GPT" is sufficient. This means humans must double-check the information provided by GPT. In fact, journalists are accustomed to and trained to reconfirm information even when it comes from reliable sources. Ultimately, the future of journalism coexisting with GPT depends solely on how wisely journalists use new technologies, and this issue applies equally to all professional fields where GPT can be utilized.

The anticipated advantages can also be summarized in a few points. For example, GPT can search, summarize, and translate large volumes of material and assist with fact-checking. It can also provide alternatives or inspiration to replace words or sentences chosen by humans, which would help improve article quality. When preparing for an interview, if GPT organizes information about the person and suggests appropriate questions based on that, journalists would have nothing to lose by referring to it at least once.

It is also useful for drafting articles or catching errors in the final stages, resembling the convenience of working alongside junior reporters, colloquially called ‘2-jin’ by journalists. The key question is whether the potential for journalism destruction is greater, and if so, journalists will surely respond by resisting GPT adoption, so this is not a problem to worry about right now.

Journalism is passive in adopting new technologies or systems and is considered one of the "slowest changing professions." How GPT can contribute to journalism’s core work and revenue models is a question difficult for individual journalists or media companies to answer. Meanwhile, if areas with purely commercial intentions seize control, it will be recorded as a second tragedy repeating the subjugation of journalism to portals. What journalists and media organizations should do now is relatively clear. We can start by warmly greeting GPT.